I’ve written previously about various aspects of the DrDoctor automated build process (you can read a few posts here, here and here). In the last week we’ve made a few more improvements that are worth a mention. The kudos for these improvements goes to my colleagues Steve (blog) and Jev.
Auto Tagging Production Releases
We have a branch that all of our releases into production go out from. We call it develop, but really we treat it more like master. Last week Steve setup TeamCity to automatically tag our git repository (hosted in BitBucket) after a successful build with the associated build number, but only on our develop (master) branch.
This is really useful, because up until that point we had no easy way of working out exactly which version of our code was running on our production box. We could work it out, but it involved clicking lots of links in Octopus Deploy, TeamCity and then git.
Now it’s the simple case of just looking at the name of the folder Octopus Deploy puts code into, as this will match the build number in TeamCity and therefore the tag in git.
HipChat Notifications on Deployment
We use HipChat as a team to communicate throughout the day, we also have TeamCity integrated with it to tell us about successful/failed builds. Now, thanks to Jev we get a notification in HipChat when a deployment has been successful/failed into either our Testing or Production environment.
This is done as the final step in our Octopus Deploy Process using a Powershell script from the Octopus Deploy library which using HipChat API broadcasts into one of our chat rooms.
We have lots of ideas in the pipeline to continue making the build and deploy process even better. Stay tuned.